Title Lack of Gender Equality Perceptions as a Predictor of Risky Sexual Behaviors among Men in Northern India

Transcription

1 Title Lack of Gender Equality Perceptions as a Predictor of Risky Sexual Behaviors among Men in Northern India Abstract Objective(s): To explore whether men s attitudes toward gender equality influence their higher-risk sexual behaviors. Design: Men s data from the north Indian states of Uttar Pradesh and Uttarakhand (n=12,240) from the National Family Health Survey (NFHS-3) were used to identify risk factors for men s reported higher-risk sexual behaviors. Key explanatory variables were based on gender equality constructs: men s attitudes to wife-beating, family violence history, views on whether women had the right to refuse having sex with their partners, and views on whether women had the right to make household decisions and have financial autonomy. Methods: Descriptive analyses were conducted to explore the relationship between higher-risk sex and each of the gender equality and socio-demographic variables. Logistic regression models were then fit to explore the independent influence of men s gender equality factors on higher-risk sex. Results: After controlling for socio-demographic factors, gender equality factors were significantly and independently associated with men s higher-risk sex. Men who were significantly more likely to engage in higher-risk sex were those who had a history of family violence [OR=1.87 ( )], viewed wife-abuse favorably [OR=1.30 ( )] and felt that women did not have the right to refuse having sex with their husbands [OR=1.52 ( )]. To reduce Indian men s higher-risk sex, HIV/STI prevention efforts must promote gender equality acceptance men, including an emphasis on the need to eliminate violence against women.

2 Introduction India s adult HIV prevalence is estimated at 0.36% of the total population, amounting to approximately 2.5 million people. This is the third largest HIV-positive population in the world, and the largest in Asia. 1 Further, studies indicate that HIV prevalence rates among high-risk populations such as injecting drug users, men who have sex with men and female sex workers range between 10-26%, and high HIV prevalence among these groups is a precursor to increased incidence among the general population. 2 Risky heterosexual sex accounts for 86% of all reported HIV infections, and 88.7% of all infections occurs in the most sexually active age group. 1 Studies examining selfreported data from India have found that up to 14.6% of all sexually active women and 4.1% of men report having at least one STI symptom, 3 and that incidence of STIs, especially syphilis, gonorrhea and type 2 herpes is increasing. 4, 5 The increased transmissibility of HIV in the presence of STIs has been well documented by numerous studies, and high prevalence of STIs is recognized as a precursor to increased HIV incidence. 6 While high HIV prevalence appeared to be confined in the eighties and nineties to southern India, 26 new districts with high HIV prevalence have been recently identified in northern India, including the states of Uttar Pradesh (UP) and Uttarakhand (which was a part of UP until 2000). 7 This is India s most populous region (with over 17% of the nation s population), lags far behind the national average in major economic and literacy indicators, and its women have less autonomy and worse health outcomes than in most other states. 8 Earlier men s studies from India have shown that higher-risk sex (sex with noncohabitating partners) takes place in special populations such as truck drivers. 9 Yet, recent studies estimate that up to 15-19% of married men and 15-47% of unmarried men in the general population might be engaging in higher-risk sexual behaviors. 10,11 Studies since the mid-nineties have suggested that HIV incidence has been increasing among married, monogamous Indian women in the general population, whose only risk factor was sexual contact with a husband who had acquired the infection primarily through higher-risk sex. 12 Studies have shown that the risk of having HIV/STIs increases for Indian women who live with abusive husbands; report that their freedom to socialize and participate in activities outside the home was being curtailed by their husbands; and 13, 14 report being concerned about husbands extramarital relationships. A number of studies emphasize that reducing men s risky sexual behaviors is key to slowing HIV 12, 14 incidence in India. Few studies have explored the link between men s gender equality attitudes and their own risk-taking behaviors. The patriarchal nature of Indian family structure instills differing sexual behavior norms for men and women; unmarried girls are expected to resist premarital sex in order to maintain their purity, and to engage in sexual activities with their husbands for procreation and motherhood. On the other hand, it is tacitly acknowledged that men can engage in higher-risk sex for the sake of gaining experience and learning to be sexual decision makers. 15, 16 Studies have reported that the reasons men give for justifying higher-risk sex include their perceived right to have access to multiple sexual partners, their perceived natural need for frequent sexual satisfaction, and married men s perceived superiority over their spouses within the 17, 18 marriage.

3 We hypothesize that men who believe that women should have less autonomy and rights than men are more likely to engage in higher-risk sexual activity. To examine the link between men s gender equality notions and their risky sexual behaviors, we used data from the National Family Health Survey (NFHS-3), the first nationally representative men s survey to collect this information in India.

4 Methods The NFHS-3 was carried out in India in two phases from November 2005 to August As part of the nationally representative survey, a total of 74,369 Indian men years were interviewed in 109,041 households. For the current analyses, data were restricted to men from the states of UP (n=11,458) and Uttarakhand (n=983). The outcome measure for this analysis was men s reported higher-risk sex, defined by the NFHS-3 as men s reported sexual intercourse with someone other than a spouse or cohabiting partner in the 12 months prior to the survey. Men reported their last three sexual partners in the 12 months prior to the survey. For this analysis, if a man reported that any one of his last three sexual partners was someone other than a spouse or cohabiting partner, that man was considered as having had higher-risk sex. The key explanatory variables in this study were men s reported measures of gender equality, assessed in terms of the following five constructs: 1. A series of seven questions in the survey asked men if they believed that a husband was justified in beating his wife under the following conditions: (1) if the husband suspected her of being unfaithful, (2) if she showed disrespect for in-laws, (3) if she went out without telling her husband, (4) if she neglected the children, (5) if she argued with him, (6) if she refused to have sex with him, and (7) if she burnt the food. Men who answered yes to any one of these were classified as believing that wife-beating was acceptable. 2. Men were asked a series of three questions on whether a woman had the right to refuse sex with her husband under the following circumstances: (1) if the husband had an STD, (2) if the husband had relations with other women, and (3) if the wife was tired or not in the mood to have sex. Men answering no to any of these questions were classified as believing that women did not have the right to refuse sex with her husband. 3. Men were asked four questions on who they thought should make the following decisions: (1) making household purchases for daily needs, (2) purchasing major household items, (3) making a decision on how many children to have, and (4) having the final say on visits to family or relatives. Men who felt that decisions should be made by women alone or jointly with their husbands were considered as believing that women should have high decision-making power, while men who felt that husbands alone should make decisions were classified as believing that women should have low decisionmaking power. 4. Men were asked who should have a final say on how a wife s earnings should be spent. Men who felt that women alone or jointly with their husbands should have this say were considered to believe that women should have financial autonomy. Men who felt that husbands alone should have the final say were considered to believe that women should not have financial autonomy. 5. Men s family violence history was measured by a single question that asked men if their fathers ever beat their mothers. Men who responded in the affirmative were considered to have a history of violence in their families. Other measures used were independent variables reported in the literature to influence higher-risk sex among men: alcohol use; mobility (whether men spent more than a month away from home in the year prior to the survey); and HIV knowledge. Men were asked if they had heard of HIV, and those who answered in the affirmative were asked the following six questions: can the risk of getting AIDS be reduced by 1) not

5 having sex at all, 2) always using condoms during sex, 3) having only one sex partner; and whether 4) a healthy person can have AIDS, 5) one can get AIDS from mosquito bites, and 6) one can get AIDS from sharing food with an infected person. A yes to questions 1-4 and no to questions 5 and 6 were deemed correct answers and given a score of one. Men scoring at the median (3) or higher were classified as having high HIV knowledge; those who score less than 3 or had not heard of HIV were classified as having low HIV knowledge. Socio-demographic measures expected to influence the outcome included men s age; urban versus rural residence; education; standard of living index; marital status; religion; caste; employment status; and whether or not a married respondent s partner lived with him. The standard of living index, represented by low, medium and high categories, was calculated by the NFHS-3 based upon ownership of household possessions, consumer durables, land and livestock. 3 Descriptive analyses were first conducted to explore the relationship between higher-risk sex and each of the gender equality and socio-demographic variables. Logistic regression models including all socio-demographic factors - deemed as important study controls were then fitted to investigate the factors that predict reported higher-risk sex. Then, secondary predictors of interest and gender equality variables were added in a stepwise manner to the models and log-likelihood tests were conducted to assess whether the addition of each of these variables helped to predict the outcome. To account for the complex survey design, the state level individual sampling weight and clustering variable (primary sampling unit) were taken into account. Statistical analyses were conducted using Intercooled Stata version 9.

6 Results Of the 12,441 men surveyed in UP and Uttarakhand states, 28 men who did not respond to the question asking about whether or not they had sex, and 173 married men who reported not having sex in the year prior to the survey, were excluded from the analysis. This resulted in a final sample size of 12,240 men. Table 1 provides self-reported prevalence of higher-risk sex among surveyed men. Of the 12,240 men in the survey who reported being sexually active in the past 12 months, 581 men (5%) reported having had higher-risk sex. Of these 581 men, 194 men (33%) used condoms during higher-risk sex. Men s socio-demographic characteristics and secondary predictors of interest are shown in Table 2. A fifth of surveyed men were in the age group of years, and nearly half of all surveyed men were between years old. Over two-thirds of men lived in a rural residence. A fourth of all men reported having no education, and most of these men were in the oldest age group of years. Nearly a quarter of surveyed men were classified as having a low standard of living, as measured by NFHS-3. Most surveyed men (93%) reported being married. Nearly two-thirds of men reported not consuming alcohol at all and 14% of men reported spending more than a month away from home in the 12 months prior to the survey. Roughly 71% of all men in the survey were found to have high HIV knowledge, as compared to 29% of men who had low or no HIV knowledge. Married and unmarried men s reported measures of gender equality are shown in Table 3. Around 40% of men indicated that wife-beating was justified, and 20% of men reported that they had witnessed their fathers physically abuse their mothers. Roughly 88% of men believed that a woman had the right to refuse having sex with her husband. Over two-thirds of all men indicated that women should have high decision-making power in the household, and 93% of men believed that a woman should have the right to spend her earnings how she wishes. The results of the final logistic regression model are shown in Table 4. After controlling for socio-demographic factors and secondary predictors of interest, at least three of the five gender equality dimensions under consideration demonstrated a statistically significant relationship with men s reported higher-risk sex. Men who believed that wife-beating was acceptable were significantly more likely to report higherrisk sex [OR=1.30 ( )], compared with men who thought wife-beating was never acceptable. Men with the higher estimated odds of having had higher-risk sex were those who believed that women did not have the right to refuse having sex with their husbands [OR=1.52 ( )]. Men who reported having a history of family violence history were significantly more likely to report higher-risk sex [OR=1.87 ( )], compared with men who reported not having a similar history of family violence. Two of the gender equality dimensions, decision-making power and men s perceptions on financial autonomy for women, were significant in bivariate analyses, but neither retained significance in the final model. Among the secondary predictors of interest, alcohol use and mobility were significantly and positively associated with men s higher-risk sex (see Table 4). Alcohol use had the largest effect size of any variable in the model, and displayed a dose-response relationship; greater frequency of alcohol use was associated with a greater likelihood of reporting higher-risk sex.

7 Discussion A small proportion of surveyed men (5%) reported having had higher-risk sex. These findings are contrary to those from smaller studies in northern India that reported roughly 15-47% prevalence of higher-risk sex among men. 10, 11 It is pointed out that strong cultural taboos regarding higher-risk sex result in under-reporting of such sexual behaviors among respondents, and studies among Indian men and women have shown that such reporting bias is far less in culturally specific interactive interviews than in faceto-face interviews such as the NFHS Despite low reported prevalence of higher-risk sex among surveyed men, these data come from the first Indian men s survey that is representative at both the national and state level. Specifically, results from this study demonstrate that among men in northern India, at least three of the five gender equality measures under consideration were found to independently associated with men s higher-risk sex. This is an important finding given that 40-50% of Indian men do not believe in at least some gender equality notions, and these men are also more likely to engage in higher-risk sex. This could provide an explanation to findings from studies in India that have found increasing HIV incidence among married monogamous Indian women. Men who felt that wife-beating was acceptable were significantly more likely to engage in higher-risk sex, as compared to men who felt that wife-beating was never acceptable. This finding has important implications for STI/HIV transmission and prevention. A number of studies have demonstrated that forced sex associated with domestic violence can cause abrasions in the mucosal lining of the vagina, thereby increasing the transmissibility of HIV and other STIs, if one of the partners is infected. 6 Studies point out that monogamous Indian women who report abuse by their husbands have higher HIV and STI prevalence rates as compared to women who are not abused. 13, 14 This study possibly provides an answer as to why this is so, by showing that men who view wife-abuse as acceptable are also more likely to engage in higher-risk sex, thereby putting themselves and their partners at increased risk of HIV/STI contraction. Family violence history was a predictor of men engaging in higher-risk sex. Previous studies have shown that men who either witnessed abuse at home or were victims of abuse in childhood are more likely in turn to be perpetrators of violence against their partners. 20 Given the links between domestic violence and increased HIV/STI transmission, and given the findings from this study that men with a history of family violence are more likely to engage in higher-risk sex, these men are at increased risk of acquiring STIs/HIV through higher-risk sex and passing on the infection to their regular partners. Prior Indian studies have found that men who are younger, more mobile, alcohol 17, 21 users, and having lesser education are more likely to engage in risky sexual activity. Results from this study confirm these findings with one exception; no significant association was observed between men s educational level and reported higher-risk sex after controlling for all other factors. Given that men who were younger, reported using alcohol, and were more mobile were highly likely to report having higher-risk sex, interventions for these groups of men remain important to curtail HIV/STI spread in 22, 23 India, as highlighted by other studies. Findings from this study have important implications for HIV/STI prevention programs in India. While it has been shown that programs aimed at increasing women s

8 empowerment improve their reproductive health outcomes, 24 altering women s autonomy in a short period of time will be difficult given the high levels of gender stratification in India and deeply entrenched patriarchal norms. 24, 25 Consequently, HIV programmatic efforts must incorporate a gender-based approach in their interventions, one that focuses on addressing men s gender equality attitudes. Studies examining men s HIV prevention interventions in high-risk-settings in India have found that sustained behavior change communication (BCC) strategies that reinforced messages of monogamy, condom use with sexual partners, and provision of government-provided condoms significantly reduced men s sex-worker visits and increased their condom use during higher-risk sex. 25 Studies recommend that existing interventions directed toward women should include partner notification and counseling services for both couples and husbands. 26, 27 Men s HIV prevention interventions should move beyond their current focus on high-risk-settings and include men in the larger community through innovative methods such as village-level peer education. HIVprevention interventions tailored specifically for men must incorporate gender-related social contexts that promote an understanding of women s autonomy and rights. 28 Given that domestic violence is a strong predictor of men s higher-risk sex, addressing this widely prevalent social problem is important not only in terms of enhancing women s rights and health, but also as an HIV prevention tool that can benefit both women and men. Such risk reducing strategies could be incorporated into alreadysuccessful men s programs aimed at reducing domestic violence in India. One such program for boys and men, MASVAW (Men s Action for Stopping Violence Against Women) is a violence-prevention campaign conducted in schools, colleges and community centers in both urban and rural settings in over forty districts of Uttar Pradesh and Uttaranchal states. 29 Following enrollment in MASVAW, men reported a qualitative improvement in their marital relationship and a reduction in violence and coercive sex with their spouses. 29 Importantly, given the findings of this study, such interventions could also reduce men s risky sexual behaviors, and hence need to be an important component of HIV prevention programs. The long term implications of such policies aimed at promoting gender equality and reducing violence against women would be to curb men s higher-risk sex and the consequent spread of HIV and other STIs in India.

9 Table 1: Self-reported higher-risk sex among men living in the north Indian states of Uttar Pradesh and Uttarakhand Higher-risk sex n (weighted %) No a 11,659 (95) Yes b 581 (5) Total 12,240 (100) a Had sexual intercourse within past 12 months exclusively with spouse or cohabiting partner b Had sexual intercourse within past 12 months with someone other than spouse or cohabiting partner

10 Table 2: Socio-demographic characteristics and secondary predictors (by weighted percentage) of men living in Uttar Pradesh and Uttarakhand Men in sample (n=12,240) Socio-demographic variables Age years years years 30 Residence Rural 72 Urban 28 Highest level of education None 25 Primary 15 Secondary 47 Secondary plus 13 Standard of living index Low 23 Medium 38 High 39 Marital status Married 93 Unmarried 7 Religion Hindu 84 Muslim 15 Other 1 Caste Scheduled caste/tribe 26 Other backward caste 47 Privileged (upper) caste 27 Employment status Employed 96 Unemployed 4 Secondary predictors Alcohol consumption Never consumed 66 Less than once a week 28 At least once a week 4 Almost daily 2 Mobility No 86 Yes 14 HIV knowledge level No or low 29 High 71

11 Table 3: Self-reported measures of gender equality (by weighted percentage) of men living in Uttar Pradesh and Uttarakhand Men in sample (n=12,240) Attitude to wife-beating Acceptable 40 Never acceptable (Ref.) 60 Family violence history Yes 20 Woman has right to refuse sex with husband No 12 Yes (Ref.) 88 Woman should have high decision-making power Yes 68 No (Ref.) 32 Woman should have financial autonomy Yes 93 No (Ref.) 7

12 Table 4: Odds ratios a and 95% confidence intervals from logistic regression models b investigating the likelihood of reporting having had higher-risk sex, among men living in Uttar Pradesh and Uttarakhand OR (95% CI) (n=12,240) Gender equality dimensions Attitude to wife-beating Acceptable 1.30* ( ) Never acceptable (Ref.) 1.0 Woman has right to refuse sex with husband No 1.52* ( ) Yes (Ref.) 1.0 Family violence history Yes 1.87* ( ) Woman should have high decision-making power Yes 1.65 ( ) Woman should have financial autonomy Yes 0.82 ( ) Secondary predictors Alcohol use Almost daily 8.72* ( ) At least once a week 4.92** ( ) Less than once a week 3.41** ( ) Never (Ref.) 1.0 Mobility Yes 1.36* ( ) Socio-demographic variables Age years 6.05* ( ) years 3.29* ( ) years (Ref.) 1.0 a Reported at p<0 001, p<0 05, and p<0 1 b Controlling for other socio-demographic factors not significant in final model: residence, education, standard of living, marital status, religion, caste, and employment status. *p<0 05, **p<0 001

13 OR (95% CI) (n=12,240) Gender equality dimensions Attitude to wife-beating Acceptable 1.30* ( ) Never acceptable (Ref.) 1.0 Woman has right to refuse sex with husband No 1.52* ( ) Yes (Ref.) 1.0 Family violence history Yes 1.87* ( ) Woman should have high decision-making power Yes 1.65 ( ) Woman should have financial autonomy Yes 0.82 ( ) Secondary predictors Alcohol use Almost daily 8.72* ( ) At least once a week 4.92** ( ) Less than once a week 3.41** ( ) Never (Ref.) 1.0 Mobility Yes 1.36* ( ) Socio-demographic variables Age years 6.05* ( ) years 3.29* ( ) years (Ref.) 1.0

14 References: 1. Indian National AIDS Control Organisation HIV Data. Available at: Accessed on January 10, Avert Who is affected by HIV and AIDS in India? Available at: Accessed on December 31, Demographic and Health Services India: DHS, 2005/ Final Report (English). Available at: Vol1[Nov ].pdf Accessed on January 10, Reynolds S, Risbud A, Shepherd M, et al. Recent Herpes Simplex Virus Type 2 Infection and the Risk of Human Immunodeficiency Virus Type 1 Acquisition in India. The Journal of Infectious Diseases. 2003;187(10): Hawkes S, Santhya KG. Diverse realities: sexually transmitted infections and HIV in India: Sexually Transmitted Infections; Coombs RW, Reichelderfer PS, Landay AL. Recent observations on HIV type-1 infection in the genital tract of men and women. Aids. Mar 2003;17(4): Indian National AIDS Control Organisation HIV Data: Breaking down the numbers. Available at: Accessed on January 10, Bloom SS, Griffiths PL. Female autonomy as a contributing factor to women's HIV-related knowledge and behaviour in three culturally contrasting states in India. Journal of Biosocial Science. 2007;39(4): Singh YN, Malaviya AN. Long distance truck drivers in India: HIV infection and their possible role in disseminating HIV into rural areas. Int J STD AIDS. Mar- Apr 1994;5(2): Bhattacharjee J, Gupta RS, Kumar A, Jain DC. Pre- and extra-marital heterosexual behaviour of an urban community in Rajasthan, India. J Commun Dis. Mar 2000;32(1): Verma RK. Sexuality in the Time of AIDS: Contemporary Perspectives from Communities in India: Sage; Newmann S, Sarin P, Kumarasamy N, et al. Marriage, monogamy, and HIV: A profile of HIV-infected women in south India. International Journal of STD & AIDS. 2000;11: Silverman JG, Decker MR, Saggurti N, Balaiah D, Raj A. Intimate partner violence and HIV infection among married Indian women. Jama. Aug ;300(6): Ball H. In India, men's sexual behavior puts their wives' reproductive health at risk. International Family Planning Perspectives. 2006;32(4): Nag M: Sexual Behavior and AIDS in India. Delhi: Vikas; 1997:xv. 16. Mukhopadadhyay S, Nandi R, Nundy M, Sivaramayya J: Gender Dimensions of HIV/AIDS: A Community Based Study in Delhi. New Delhi, Institute of Social Studies Trust, Schensul SL, Mekki-Berrada A, Nastasi BK, Singh R, Burleson JA, Bojko M. Men's extramarital sex, marital relationships and sexual risk in urban poor communities in India. Journal of Urban Health-Bulletin of the New York Academy of Medicine. 2006;83(4):

15 18. Verma RK, Mahendra VS, Pulerwitz J, Barker GT, Van Dam J, Flessenkaemper S. From research to action: Addressing masculinity and gender norms. Indian Journal of Social Work. Oct 2004;65(4): Jaya, Hindin MJ, Ahmed S. Differences in young people's reports of sexual behaviors according to interview methodology: a randomized trial in India. Am J Public Health. Jan 2008;98(1): Fonagy P. Male Perpetrators of Violence Against Women: An Attachment Theory Perspective. Journal of Applied Psychoanalytic Studies. 1999;1(1): Vaz FS, Ferreira AM, Kulkarni MS, Motghare DD. Sexual risk behaviors and HIV/AIDS awareness among males in a rural community in Goa. J Commun Dis. Mar 2006;38(1): Sivaram S, Srikrishnan AK, Latkin C, et al. Male alcohol use and unprotected sex with non-regular partners: Evidence from wine shops in Chennai, India. Drug Alcohol Depend. Jan Chandrasekaran P, Dallabetta G, Loo V, Rao S, Gayle H, Alexander A. Containing HIV/AIDS in India: the unfinished agenda. Lancet Infectious Diseases. 2006;6(8): Bloom SS, Wypij D, Das Gupta M. Dimensions of women's autonomy and the influence on maternal health care utilization in a North Indian city. Demography. Feb 2001;38(1): Bentley ME, Spratt K, Shepherd ME, et al. HIV testing and counseling among men attending sexually transmitted disease clinics in Pune, India: changes in condom use and sexual behavior over time. Aids. 1998;12(14): Pallikadavath S, Jayachandran AA, Stones RW. Women s Reproductive Health, Sociocultural Context and AIDS Knowledge in Northern India. Journal of Health Management. 2005;7(1): Gangakhedkar RR, Bentley ME, Divekar AD, et al. Spread of HIV infection in married monogamous women in India. Jama-Journal of the American Medical Association. Dec 1997;278(23): Go VF, Sethulakshmi CJ, Bentley ME, et al. When HIV-prevention messages and gender norms clash: The impact of domestic violence on women's HIV risk in slums of Chennai, India. Aids and Behavior. Sep 2003;7(3): Men s Action for Stopping Violence against Women. Save the Children Available at: 0aa685a9a68d/MASVAW+Final+Report+11+June+2008.pdf?view=1&part=4 Accessed on January 10, 2009.